Project Summary/Abstract The broad, long-term, scientific objective of this project is to identify the behavioral and neural factors that determine the efficacy of robotic hand movement training after stroke. An important societal impact of achieving this objective will be the design of more effective robotic rehabilitation exercise technology, which will allow people with a stroke to increase their movement recovery beyond that possible with current approaches. In our previous grant, we found that participants with impaired finger proprioception (quantified robotically with a novel protocol) did not achieve a functional benefit from robotic finger training. Reduced benefit also correlated with injury to and abnormal activation of the somatosensory system. Our working hypothesis is that finger proprioceptive integrity is a gateway for robotic assistance because it allows such assistance to stimulate a Hebbian-like learning mechanism. Building on this ?Hebbian Hypothesis?, we add two other hypotheses: improving proprioceptive capacity with targeted training for those with impaired proprioception will enhance the effectiveness of subsequent robotic finger training, and proprioceptive integrity modulates spontaneous self- training outside of formal training. We will test these hypotheses in a series of experiments with a novel robotic finger training device (?FINGER?) that assists participants in making different grips to play notes in a musical computer game similar to Guitar Hero. We will recruit participants with hand movement deficits at least six months after stroke and randomize the participants with intact finger proprioception to participate in Aim 1, and those with impaired finger proprioception to participate in Aim 2. Aim 1 is to identify the magnitude of the Hebbian benefit from robot-assisted movement training. Robotic assistance enhances proprioceptive input, but it typically also enhances reward, because it increases task success. We will determine the magnitude of the benefit of the enhanced proprioceptive input, beyond the benefit due to enhanced reward. Aim 2 is to determine the extent to which finger proprioception can be improved through targeted robotic proprioceptive training, and thereby enhance response to subsequent robotic finger movement training. Aim 3 is to identify mathematical models that predict the response to the different forms of training experienced by the participants in Aim 1 and 2. The model inputs will be baseline behavioral and demographic measures, lesion overlap with sensory and motor structures (via anatomical MRI), and activation of sensory-motor networks (via fMRI and EEG). We aim to provide further insight into the high variance in robotic training response, and determine who responds best to proprioceptive training. Aim 4 is to identify the role of impaired proprioception in decreasing spontaneous hand use in daily life using a novel wearable sensor, because hand use outside of robotic therapy sessions likely exerts a powerful training effect. Testing of these hypotheses will provide insight into how to manipulate the unique resources for sensory motor plasticity that each patient retains using robot-based movement training.